This paper uses an updated and revised panel data set on ambient air pollution in cities worldwide to examine the robustness of the evidence for the existence of an inverted U-shaped relationship between national income and pollution. We test the sensitivity of the pollution-income relationship to functional forms, to additional covariates, and to changes in the nations, cities, and years sampled. We find that the results are highly sensitive to these changes, and conclude that there is little empirical support for an inverted U-shaped relationship between several important air pollutants and national income in these data.